Plot temperature vs humidity

Load required packages:

require(data.table)
require(ggplot2)
require(plotly)
require(RSQLite)

Fetch data from db:

# set the file path to db
s.dbpath = "../data/dhtlog.db"
# connect to db
con <- dbConnect(drv=RSQLite::SQLite(), dbname=s.dbpath)
# list all tables
tables <- dbListTables(con)
# exclude sqlite_sequence (contains table information)
tables <- tables[tables != "sqlite_sequence"]
lDataFrames <- vector("list", length=length(tables))
# create a data.frame for each table
for (i in seq(along=tables)) {
  lDataFrames[[i]] <- as.data.table(dbGetQuery(conn=con, statement=paste("SELECT * FROM '", tables[[i]], "'", sep="")))
}
# take 1st table only (this is where our data lives)
dt = lDataFrames[[1]]

Get season from dates (adapted from https://stackoverflow.com/a/9501225/1898713):

getSeason <- function(DATES) {
  WS <- as.Date("2012-12-21", format = "%Y-%m-%d") # Winter Solstice
  SE <- as.Date("2012-3-20",  format = "%Y-%m-%d") # Spring Equinox
  SS <- as.Date("2012-6-21",  format = "%Y-%m-%d") # Summer Solstice
  FE <- as.Date("2012-9-22",  format = "%Y-%m-%d") # Fall Equinox
  
  # Convert dates from any year to 2012 dates
  d <- as.Date(strftime(DATES, format="2012-%m-%d"))
  
  ifelse (d >= WS | d < SE, "Winter",
          ifelse (d >= SE & d < SS, "Spring",
                  ifelse (d >= SS & d < FE, "Summer", "Fall")))
}
# add a column 'season' to our data
dt[, season := getSeason(timestamp)]
# order season column
dt[, season := factor(season, levels = c('Winter', 'Spring', 'Summer', 'Fall'))]

Sometimes sensors malfunction and give strange readings such as very low temperature or very high humidity. Remove such measurements:

dt = dt[temp > 10 & humid < 100]

Plot interactive scatter plot with points coloured by season. To avoid slowdown in plotly there are nPts random points selected and plotted for each season.

# number of points to select
nPts = 500
# create a new trimmed dt with nPts random points from each season
dt.short = dt[, .SD[sample(.N, min(.N, nPts))], by = season]
# create a ggplot
p1 = ggplot(dt.short, aes(x = temp, y = humid, label = timestamp)) +
  geom_point(alpha = 0.3, aes(fill = season), colour = NA) +
  xlab('\nTemperature') + 
  ylab('Humidity\n') +
  theme_bw()
# interactive scatter plot
ggplotly(p1)